Published online September 16, 2005
Nucleic Acids Research, 2005, Vol. 33, No. 16 5297–5307
doi:10.1093/nar/gki823
Stabilizing contributions of sulfur-modified
nucleotides: crystal structure of a DNA duplex
with 20-O-[2-(methoxy)ethyl]-2-thiothymidines
Benjamin Diop-Frimpong, Thazha P. Prakash1, Kallanthottathil G. Rajeev2,
Muthiah Manoharan2 and Martin Egli3,*
Department of Chemical Engineering, School of Engineering, Vanderbilt University, Nashville, TN 37235, USA,
1
Department of Medicinal Chemistry, Isis Pharmaceuticals Inc., Carlsbad, CA 92008, USA, 2Drug Discovery,
Alnylam Pharmaceuticals Inc., Cambridge, MA 02142 USA and 3Department of Biochemistry,
School of Medicine, Vanderbilt University, Nashville, TN 37232, USA
Received May 27, 2005; Revised and Accepted August 23, 2005
ABSTRACT
Substitution of oxygen atoms by sulfur at various
locations in the nucleic acid framework has led
to analogs such as the DNA phosphorothioates and
40 -thio RNA. The phosphorothioates are excellent
mimics of DNA, exhibit increased resistance to nuclease degradation compared with the natural counterpart, and have been widely used as first-generation
antisense nucleic acid analogs for applications in vitro
and in vivo. The 40 -thio RNA analog exhibits significantly enhanced RNA affinity compared with RNA, and
shows potential for incorporation into siRNAs. 2Thiouridine
(s2U)
and
5-methyl-2-thiouridine
5 2
(m s U) are natural nucleotide analogs. s2U in tRNA
confers greater specificity of codon–anticodon interactions by discriminating more strongly between
A and G compared with U. 2-Thio modification preorganizes the ribose and 20 -deoxyribose sugars for
a C30 -endo conformation, and stabilizes heteroduplexes composed of modified DNA and complementary RNA. Combination of the 2-thio and sugar
20 -O-modifications has been demonstrated to boost
both thermodynamic stability and nuclease resistance. Using the 20 -O-[2-(methoxy)ethyl]-2-thiothymidine (m5s2Umoe) analog, we have investigated the
consequences of the replacement of the 2-oxygen
by sulfur for base-pair geometry and duplex
conformation. The crystal structure of the A-form
DNA duplex with sequence GCGTAT*ACGC
(T* = m5s2Umoe) was determined at high resolution
and compared with the structure of the corresponding
duplex with T* = m5Umoe. Notable changes as a result
of the incorporation of sulfur concern the base-pair
parameter ‘opening’, an improvement of stacking in
the vicinity of modified nucleotides as measured by
base overlap, and a van der Waals interaction between
sulfur atoms from adjacent m5s2Umoe residues in
the minor groove. The structural data indicate only
minor adjustments in the water structure as a result
of the presence of sulfur. The observed small structural perturbations combined with the favorable consequences for pairing stability and nuclease
resistance (when combined with 20 -O-modification)
render 2-thiouracil-modified RNA a promising
candidate for applications in RNAi.
INTRODUCTION
Chemically modified nucleic acids have been studied extensively in the context of the development of antisense therapeutics (1,2) and more recently in the search for small interfering
RNAs (siRNAs) for in vitro and in vivo applications (3,4). In
both cases, increases in RNA affinity as a result of modification can be expected to play an important role in the improvement of the efficacy of putative oligonucleotide therapeutics
(5). Phosphorothioate DNA (PS-DNA) (6) has undergone
extensive tests in numerous clinical trials of antisense oligonucleotides (7). In PS-DNA, one of the non-bridging phosphate oxygen atoms is replaced by sulfur. Although this
altered chemistry leads to a slight reduction in RNA affinity
relative to DNA, PS-DNA was considered a promising firstgeneration antisense modification for a number of reasons,
including ease of synthesis, increased nuclease resistance
*To whom correspondence should be addressed. Tel: +1 615 343 8070; Fax: +1 615 322 7122; Email: [email protected]
The Author 2005. Published by Oxford University Press. All rights reserved.
The online version of this article has been published under an open access model. Users are entitled to use, reproduce, disseminate, or display the open access
version of this article for non-commercial purposes provided that: the original authorship is properly and fully attributed; the Journal and Oxford University Press
are attributed as the original place of publication with the correct citation details given; if an article is subsequently reproduced or disseminated not in its entirety but
only in part or as a derivative work this must be clearly indicated. For commercial re-use, please contact [email protected]
5298
Nucleic Acids Research, 2005, Vol. 33, No. 16
and binding to proteins (biodistribution) as well as degradation
of RNA bound to PS-DNA by RNase H (8). Another sulfur
modification, 2-thiouridine (s2U) was found to strongly
increase the melting temperature (Tm) of heteroduplexes
between modified strands and RNA compared with the
corresponding DNA:RNA hybrids (9). For example, incorporation of a single s2U residue in the center of a nonamer
DNA strand led to an increase of 9.4 C in the Tm for the
duplex with the complementary RNA strand compared with
the unmodified hybrid duplex (10). Consecutive and
interspersed replacement of several thymidines by 20 -O-[2(methoxy)ethyl]-2-thiothymidine (m5s2Umoe; Figure 1) furnished an average gain in Tm of 3.5 and 3 C, respectively, for
heteroduplexes between modified strands and RNA relative to
the native duplexes (11).
The s2U modification occurs naturally in transfer RNAs
[(12–14); see also the RNA modification database http://
medstat.med.utah.edu/RNAmods/]. The modification is present at the first position of the anticodon, position 34, and
affects the relative stabilities of pairing with A and the wobble
pairing with G of codons, thereby stabilizing and increasing
the specificity of codon–anticodon interactions. Because sulfur
Figure 1. Structure of 20 -O-[2-(methoxy)ethyl]-2-thiothymidine (s2m5Umoe).
has a bigger atomic radius and lower electronegativity than
oxygen, its H-bonding ability is weaker relative to the latter.
As a result, the interaction between the sulfur in the 2-position
of s2U with the hydrogen on N1 of G (Figure 2A, right) is
weakened compared with the corresponding pair between U
and G (Figure 2B, right). On comparison, the hydrogen bond
between N1 of A and N3-H of s2U (Figure 2A, left) is only
indirectly affected, rendering its stability similar to that of
the U:A pair (Figure 2B, left). In summary, in the absence
of the sulfur modification, the more electronegative oxygen in
the 2-position of U will readily form a hydrogen bond with the
hydrogen on N1 of G. This makes the U:G interaction more
similar to the U:A interaction (Figure 2B) compared with the
relative stability of the s2U:G and s2U:A pairs (Figure 2A)
[(15) and cited references]. In addition to conferring pairing
specificity, the s2U modification also restricts uridine
dynamics by locking the nucleoside in the anti, C30 -endo
conformation (16,17). Moreover, the conformational effects
of the modification extend into the 30 -adjacent nucleosides,
causing them to assume a similar conformation (18,19). The
strong conformational preorganization towards the C30 -endo
pucker by the sulfur at the 2-position of the base is consistent
with the aforementioned increased RNA affinity exhibited by
oligonucleotides containing the s2U modification.
To examine the potential conformational changes in a
duplex as a consequence of replacement of T by m5s2U
and to better understand the origins of the enhanced stability
of s2U-modified duplexes compared with their native
counterparts, we determined the X-ray crystal structure of
an A-form DNA decamer duplex with sequence 50 GCGTA(m5s2Umoe)ACGC and two 20 -O-[2-(methoxy)ethyl]-2-thiothymidine residues. The structure of the duplex
with the same sequence containing a single 20 -O-[2-(methoxy)ethyl]-thymidine per strand (50 -GCGTA(m5Umoe)ACGC)
had previously been determined at a similar resolution and
was used as the reference (20). The 20 -O-MOE RNA analog is
Figure 2. Pairing of U and s2U with A and G. (A) Diagram illustrating s2U base-pairing with both A and G; s2U favors A over G. (B) Diagram illustrating the lack in
specificity with uridine; U binds with similar affinity to both A and G.
Nucleic Acids Research, 2005, Vol. 33, No. 16
a promising second-generation antisense modification that
leads to significant gains in RNA affinity as well as enhanced
nuclease resistance compared with DNA or PS-DNA (21,22).
The roles of conformational preorganization, stereoelectronic
effects and hydration in the stability increases afforded by the
20 -O-MOE modification have been studied extensively using
crystallographic data obtained for partially (20) and all-20 -Omodified duplexes (23).
Here, we show that the presence of sulfur at the 2-position of
T results in subtle changes in the local helical parameters and
goes along with a slight improvement of the stacking interactions with adjacent bases, evident from the comparison of
the orientations of the m5s2U and m5U bases relative to their
nearest neighbors in the m5s2Umoe-modified and reference
duplex, respectively. Overall the structural data are consistent
with the significant gains in thermodynamic stability afforded
by the s2U modification.
MATERIALS AND METHODS
5299
Table 1. Selected crystal data, data collection and refinement statistics
Crystal data
Space group
a (Å)
b (Å)
c (Å)
Data collection
Source/detector
Temperature ( C)
Wavelength (Å)
Total no. of reflections
No. of unique reflections
Resolution range (Å)
Completeness, all/1.71–1.61 Å (%)
R-merge, all/1.71–1.61 Å (%)
Refinement statistics
No. of DNA atoms
No. of water molecules
R.m.s. distances (Å)
R.m.s. angles ( )
Mean B value (overall, Å2)
R-work
R-free
P212121
25.25
45.00
44.97
DND-CAT 5-ID (APS)/MAR225
170
0.9873
51 837
6304
32.0–1.61
95.0/88.2
7.1/15.5
460
160
0.01
2.4
9.1
0.195
0.202
Synthesis of the 2-thio-20 -O-MOE-5-methyluridine
modified oligonucleotide
The
50 -O-DMT-20 -O-MOE-2-thiothymidine-30 -phosphoramidite was prepared according to the published procedure
(11). The oligonucleotide d(GCGTA)-m5s2Umoe-d(ACGC)
was synthesized on functionalized controlled pore glass
(CPG) on an automated solid phase DNA synthesizer with
final DMT group retained at the 50 end. Standard phosphoramidites and solid supports were used for incorporation of A,
T, G and C residues. A 0.1 M solution of the amidites in
anhydrous acetonitrile was used for the synthesis of the modified oligonucleotide. For incorporation of 2-thio-20 -O-MOE-5methyl U amidite, 6 equivalents of phosphoramidite solution
were delivered in two portions, each followed by a 5 min
coupling wait time. All other steps in the protocol supplied
by the manufacturer were used without modification. Oxidation of the internucleotide phosphite to the phosphate was
carried out using tert-butyl hydroperoxide/acetonitrile/water
(10:87:3) with 10 min oxidation wait time. The stepwise coupling efficiencies were >97%. After completion of the synthesis,
the solid support was suspended in aqueous ammonium
hydroxide (30 wt%) and kept at room temperature for 2 h.
The solid support was filtered and the filtrate was heated at
55 C for 6 h to complete the removal of all protecting groups.
The crude oligonucleotide was purified by high performance
liquid chromatography (HPLC, C-4 column, Waters,
7.8 · 300 mm, 15 mm, 300 Å, buffer A ¼ 100 mM ammonium acetate, pH 6.5–7, buffer B ¼ acetonitrile, 5–60% of B
in 55 min, flow 2.5 ml min1, l 260 nm). Detritylation was
achieved by adjusting the pH of the solution to 3.8 with acetic
acid and keeping at room temperature until complete removal
of the trityl group, as monitored by HPLC analysis. The
oligonucleotide was then desalted by HPLC and characterized
by ES-MS (calculated mass: 3118.35 g mol1; found:
3119.57 g mol1) and purity (91% full length) was assessed
by capillary gel electrophoresis.
Crystallizations and data collection
Crystallization trials were performed with the Nucleic Acid
Mini-screen (24) by Hampton Research (Aliso Viejo, CA),
using the hanging drop vapor diffusion technique. Droplets
with volume 2 ml of a 1:1 mixture of sample and mini-screen
buffer were equilibrated against 1 ml of 35% 2-methyl-2,4pentanediol (MPD). Crystals were obtained from a droplet that
contained 1 mM oligonucleotide, 40 mM sodium cacodylate
pH 7.0, 12 mM spermine tetrahydrochloride, 12 mM sodium
chloride, 80 mM potassium chloride and 10% v/v MPD. A
single crystal was mounted in a nylon loop and frozen in liquid
nitrogen. Diffraction data were collected at low temperature in
a cold nitrogen stream on the 5-ID beamline of the DuPontNorthwestern-Dow Collaborative Access Team at the
Advanced Photon Source, Argonne, IL. Separate data sets
for high and low resolution reflections were acquired. All
data were processed with the program XDS (25) and a summary of selected crystal data and data collection statistics is
listed in Table 1.
Crystal structure determination and refinement
Due to the close similarity between two of the unit cell constants, the diffraction data were processed both in the space
groups P222 (orthorhombic) and P422 (tetragonal). The resulting values for R-merge differed only minimally (<1%) and
structure determination was subsequently attempted in
P212121, the space group found for the vast majority of crystals
of A-form DNA 10mers with isolated 20 -O-modified residues
investigated in our laboratory (20), and in space group P42212.
Phasing was carried out by the molecular replacement method
using the program AMORE (26) as part of the CCP4 suite of
crystallographic programs (27). An A-form DNA decamer of
the same sequence GCGTATAGCG (28) was used as the
search model. Promising solutions with good R-factors, reasonable correlation coefficients, and few short lattice contacts
were found in both space groups. Initial refinements of the
models were performed with the program CNS (29), setting
aside 5% randomly selected reflections for calculating the
R-free (30). While the R-factor calculated for the model in the
orthorhombic space group P212121 dropped rapidly, it failed
5300
Nucleic Acids Research, 2005, Vol. 33, No. 16
Figure 3. Example of the final sum (2Fo-Fc) Fourier electron density (1s threshold) around the base-pair step (A5)p(m5s2Umoe6):(A15)p(m5s2Umoe16). The view is
into the minor groove with 20 -O-MOE substituents in the foreground (m5s2Umoe6, left, and m5s2Umoe16, right), and atoms are colored green, blue, red, orange and
yellow for carbon, nitrogen, oxygen, phosphorus and sulfur, respectively.
to drop below 35% in space group P42212. Consequently,
refinement was continued only in P212121. After rigid-body
refinement and simulated annealing plus multiple rounds of
positional and individual B-factor refinement, the complete
model, including the 2-thio and 20 -O-MOE modifications
for the two modified residues, exhibited an R-factor of
30.1% (R-free of 28.5%). Refinement was then continued
with the progam REFMAC (31), using a bulk solvent correction and treating DNA atoms and water molecules with
restraint anisotopic B-factors (32,33). The final model refined
to an R-factor of 19.4%, using all reflections up to a resolution
of 1.61 Å. Selected refinement parameters are listed in Table 1
and an example of the quality of the final Fourier sum (2Fo-Fc)
electron density is depicted in Figure 3.
Coordinates
Final coordinates and structure factors have been deposited in
the Protein Data Bank (http://www.rcsb.com) PDB ID code
2AXB.
RESULTS
Overall structure
Owing to the similar b and c unit cell constants (Table 1),
diffraction data from a single crystal of the m5s2Umoe-modified DNA decamer were indexed and processed both in orthorhombic and tetragonal space groups. Following structure
determination with molecular replacement using an A-form
model, initial rounds of refinement in the space groups
P212121 and P42212 indicated that the crystal system is orthorhombic. Thus, the crystallographic asymmetric unit consists
of a single duplex and the lattice features a noncrystallographic 4-fold symmetry. Multiple rounds of coordinate and temperature factor refinements and simulated
annealing led to an improved model, and electron density
maps based on it revealed ordered 20 -O-MOE substituents
in the minor groove. The positions of sulfur atoms were
marked by spherical peaks of density that are clearly enlarged
relative to those around the exocyclic O4 and O2 atoms of
unmodified thymines (Figure 3). In the duplex, residues of one
strand are numbered from 1 to 10, and those in the complementary strand are numbered from 11 to 20; m5s2Umoe
residues are located at positions 6 and 16.
The modified duplex adopts an A-form conformation with
an average helical rise of 2.64 Å (SD 0.69 Å) an average
helical twist of 34.3 (SD 3.4 ). The average inclination of
base pairs relative to the helical axis is 15.6 . Except for
residues at the 50 -terminal ends, the conformations of all sugars fall into the C30 -endo range. The deoxyriboses of residues
G1 and C2 adopt C20 -exo and C40 -exo puckers, respectively,
and G11 is flipped into the C20 -endo conformation. Similarly,
most of the backbone torsion angle conformations observed in
the decamer duplex are those typically associated with an
A-form geometry (-sc/ap/sc/sc/ap/-sc; a to z). Exceptions
are constituted by A5 in strand 1 and G13 in strand 2 that
both assume an extended backbone variant with a, b and g in
the ap conformation. The conformation is brought about by a
crankshaft motion of the backbone and is accompanied by
elongated distances between adjacent intra-strand phosphorus
atoms (6.9 Å between A5 and m5s2Umoe6 and 6.4 Å between
G13 and T14; avg P. . .P distance: 6.1 Å).
To analyze the geometry of the m5s2Umoe-modified duplex
further and to determine potential changes as a result of the
presence of sulfur atoms in the central A:m5s2Umoe base
pairs, we used the previously determined structure of the
duplex with identical sequence and m5Umoe residues at positions 6 and 16 as a reference (20). The resolution of that
structure is similar (1.93 Å) to that of the m5s2Umoe-modified
duplex described here. Both crystals belong to space group
P212121 and have very similar cell constants [a ¼ 24.93,
b ¼ 44.59, c ¼ 45.38 Å (reference) versus a ¼ 25.25,
b ¼ 45.00, c ¼ 44.97 Å]. We will refer to the reference
Nucleic Acids Research, 2005, Vol. 33, No. 16
and the 2-thio-modified duplexes as the m5Umoe and
m5s2Umoe structures, respectively.
Crystal packing effects
Because the conformational perturbations caused by the 2-thio
modification can be expected to be quite subtle, it is important
to identify and quantify conformational differences between
the m5s2Umoe-modified and reference duplexes that are due to
crystal packing effects. Interactions between neighboring
duplexes in the crystal lattices of m5Umoe and m5s2Umoe
decamers involve stacking of terminal base pairs into the
outer portions of the minor groove (34). This particular
mode of interaction triggers various degrees of kinking into
the major groove. Individual crystallization conditions, the
chemistry and spatial properties of substituents of 20 -Omodified residues and their location in the decamer strand,
as well as data collection temperature all appear to play a
role in the relative degree of kinking (35). Superimposition
of the two duplexes reveals that the m5Umoe decamer exhibits
a slightly stronger kink than the m5s2Umoe decamer (Figure 4).
5301
The kink compresses the major groove and occurs between the
tetramer 50 -G1pC2pG3pT4 and the hexamer pA5p(m5Umoe6)pA7pC8pG9pC10. It is accompanied by significant
changes in the local base-pair (stagger), base-pair step (rise,
twist and roll) and base-pair helical parameters (inclination).
The absolute values of these parameters at that site constitute
either the minimum (stagger [m5Umoe], rise, twist) or maximum (stagger [m5s2Umoe], roll, inclination) among the individual observations with both duplexes. Most probably, these
changes are only weakly affected by the oxygen!sulfur substitution in residue 16, although the kink occurs between the
T4:A17 and A5:m5s2Umoe16 pairs. Rather, they are a reflection of the intrinsic conformational properties (‘bendability’)
of the decamer sequence and the packing forces in the orthorhombic crystal lattice. Regarding the possible origins of the
observed difference in kinking in the two duplexes, it is noteworthy that the m5Umoe structure was determined at room
temperature (20) whereas the m5s2Umoe structure is derived
from diffraction data collected at 170 C. Thus, slightly
divergent packing forces due to the temperature change are
the most likely cause of the observed differences between the
global conformations of the m5s2Umoe and m5Umoe
duplexes. Because of the kink adjacent to the A5:m5s2Umoe16
base pair and the ensuing perturbations in its immediate vicinity, one would expect that careful analysis of the geometry of
the other modified base pair, m5s2Umoe6:A15, might yield a
more reliable estimate of the changes as a result of the 2-thio
modification.
Effects of the 2-thio modification on the
local conformation
Figure 4. Overall conformation of the m5s2Umoe-modified decamer (atoms
are colored green, blue, red, orange and yellow for carbon, nitrogen, oxygen,
phosphorus and sulfur, respectively) compared with the m5Umoe-modified
reference duplex (thin blue lines). The view of the A-form duplexes is across
the major (top) and minor grooves (bottom) and the 50 -terminal G1 and G11
residues are located at the top left and bottom left, respectively. 20 -O-Substituents are jutting into the minor groove and 2-sulfur atoms highlighted in
yellow. The reduced kink into the major groove in the m5s2Umoe-modified
decamer between the tetramer 50 -G1pC2pG3pT4:30 -C20pG19pC18pA17p and
the hexamer 50 -pA5p(m5s2Umoe6)pA7pC8pG9pC10:30 -(m5s2Umoe16)pA15pT14pG13pC12pG11 is clearly visible.
Both structures were analyzed with the program 3DNA that
allows calculation of geometric parameters for nucleic acid
molecules (36). Among the parameters, one would expect
those that concern the local geometry of base pairs and the
stacking interactions to be the most informative in terms of
potential conformational changes due to 2-thio modification.
Figure 5 depicts graphic representations of selected geometric
parameters for base pairs in the m5s2Umoe and reference
duplexes. A significant change (11 ) is observed in the parameter opening for base pair m5s2Umoe6:A15 compared with
the corresponding pair in the reference structure (Figure 5A).
This change in opening results in a larger separation
between the C2 atoms of the two bases in the minor
groove with the 2-thio modified pair [4.23 Å versus 3.89 Å
(ref. structure); Figure 6A]. Comparison of the stretch parameters also indicates a small change (0.15 Å; not shown)
for this base pair that is related to accommodating the
larger sulfur atom. Compared with the above change in opening, the differences in the parameters shear, stagger and propeller twist (Figure 5B–D, respectively) are relatively minor
for base pair m5s2Umoe6:A15. Likewise, it does not appear
that buckling of bases is affected by the sulfur substitution.
The calculated values for both the m5s2Umoe6:A15 and the
A5:m5s2Umoe16 pairs differ by <1 from those for the
corresponding pairs in the reference structure. Interestingly,
the latter base pair does not exhibit an altered opening angle
relative to the reference duplex (Figure 5A). The difference
amounts to less, 2 although the separation between
C2 atoms of bases (4.11 Å) is comparable with that seen in
5302
Nucleic Acids Research, 2005, Vol. 33, No. 16
A
B
10
Opening (deg.)
6
0.6
4
0.4
2
0.2
0
0
-2
G
C
G
T
A
s2U
A
C
G
C
-0.2
-4
-0.4
-6
-0.6
C
0.8
0.6
0.4
0.2
0
-0.2
-0.4
-0.6
-0.8
-1
-1.2
-1.4
G
C
G
T
A
s2U
A
C
G
C
s2U
A
C
G
C
Aps2U s2UpA
ApC
CpG
D
Stagger (Å)
Propeller twist (deg)
15
10
5
0
-5
G
C
G
T
A
2
sU
A
C
G
G
C
G
T
A
-10
C
-15
-20
-25
-30
E
F
x-displacement (Å)
y-displacement (Å)
3.5
3
2.5
2
1.5
1
0.5
0
-0.5
-1
-1.5
-2
-2.5
-3
Shear (Å)
0.8
8
0
-1
GpC
CpG
GpT
TpA
GpC
-2
-3
-4
-5
GpC
CpG
GpT
TpA
Aps2U s2UpA ApC
CpG
GpC
-6
-7
-8
Figure 5. Local inter-base parameters (A) ‘opening’, (B) ‘shear’, (C) ‘stagger’ and (D) ‘propeller twist’, and local base-pair helical parameters (E) ‘y-displacement’
and (F) ‘x-displacement’ in the m5s2Umoe- (filled circles, thin solid lines) and m5Umoe-modified decamer duplexes (filled squares, dashed lines). All parameters
were calculated with the program 3DNA (36) and a cartoon of the particular parameter is shown at the upper right in each panel.
the m5s2Umoe6:A15 pair (see above). Similarly, other
parameters, including stretch, shear, stagger and propeller
twist change subtly, but none of them appears to be making
the major contribution such as opening in case of the
m5s2Umoe6:A15 pair (Figure 5). Rather, the combination of
many small changes results in an arrangement that shifts
the sulfur atom away from the C2 atom of adenine in the
minor groove. In any case, a superposition of the
s2m5Umoe6:A15 and A5:s2m5Umoe16 pairs demonstrates
that they adopt relatively similar conformations, with virtually
identical relative orientations of C2(A) and S2(m5s2Umoe)
(Figure 6B).
In order to examine whether presence of the sulfur atoms in
the m5s2Umoe structure led to changes in the stacking interactions involving the four central base pairs, we calculated the
overlap areas between adjacent pairs in Å2 units (36). The
3DNA program allows computation of the surface areas of
polygons projected in the mean plane of a base pair at a
particular step, including exocyclic ring atoms. Stacking interactions at selected base-pair steps are shown in Figure 7.
Nucleic Acids Research, 2005, Vol. 33, No. 16
A
B
Figure 6. Effects of the 2-thio modification on the geometry of m5s2Umoe:A
base pairs. (A) Superimposition of the m5s2Umoe6:A15 (colored by atom) and
m5Umoe6:A15 (black lines) base pairs, illustrating base opening in the m5s2Umoe:A base pair. (B) Superimposition of the A5:m5s2Umoe16 (colored by
atom) and m5s2Umoe6:A15 (black lines) base pairs, illustrating the nearly
identical orientations of the C2(A) and S2(m5s2Umoe) atoms in the two pairs.
The views are approximately along the normal to the base planes and Watson–
Crick hydrogen bonds are drawn as thin solid lines.
Inspection of the plotted diagrams for the m5s2Umoe and
reference duplexes reveals minor alterations in stacking at
most base-pair steps. An exception is found at the
(A5)p(m5s2Umoe6):(A15)p(m5s2Umoe16) step (Figure 7C)
that displays a clearly visible shift between base pairs relative
to the reference structure. Indeed, the calculated y-displacement at that site shows a difference of 2.3 Å between the
m5s2Umoe and m5Umoe structures (1.1 versus 1.2 Å,
respectively; Figure 5E). The difference in x-displacement
is somewhat smaller, 1.3 Å (5.3 versus 4.0 Å in
m5s2Umoe and m5Umoe, respectively; Figure 5F), but constitutes the largest change for that parameter in the entire duplex.
The two neighboring steps also show considerable changes in
y-displacement that amount to 1.6 Å (Figure 7B) and 1.7 Å
(Figure 7D). At the central step, it is readily apparent that
the two sulfur atoms have moved toward one another. Indeed,
the distance between the sulfur atoms (3.48 Å) is shorter
than the one between the corresponding O2 atoms in
the reference structure (3.73 Å). The former is slightly shorter
than the sum of van der Waals radii for the two sulfur
atoms (3.6 Å). When the overlap areas between adjacent
base pairs at the (T4)p(A5):(m5s2Umoe16)p(A17),
(A5)p(m5s2Umoe6):(A15)p(m5s2Umoe16) and (m5s2Umoe6)
p(A7):(T14)p(A15) steps (Figure 7B–D) are added, a slight
increase in stacking for the m5s2Umoe decamer compared with
the reference structure is apparent (17 Å2 versus 15 Å2). Therefore, we can conclude that the experimental data regarding
stacking are at least in line with the stabilizing effect of
the 2-thio modification.
5303
Hydration of sulfur atoms and MOE moieties
In the minor groove, sulfur atoms of adjacent m5s2Umoe residues are bridged by a single water molecule. This water, in
turn, forms a hydrogen bond to the outer oxygen of the 20 -OMOE substituent of m5s2Umoe16, and a second, relatively
weak hydrogen bond to the 40 -oxygen of the same residue.
The 2-sulfur atom of m5s2Umoe6 is hydrogen bonded to a
second water molecule. This water bridges the nucleobase
to a further water molecule that is itself hydrogen bonded
to both O20 and the outer oxygen of the 20 -O-MOE substituent
from m5s2Umoe6. Both 20 -O-MOE moieties adopt gauche
conformations (m5s2Umoe6, sc+, and m5s2Umoe16, sc,
Figure 3). In the case of m5s2Umoe6, the substituent is directed
toward the center of the minor groove, allowing for a direct
link between the MOE moiety and the minor groove edge of
the base by a single water molecule. On comparison, the
substituent of m5s2Umoe16 is directed away from the base
edge, therefore requiring a water tandem to link MOE moiety
and S2. Water molecules trapped between the 20 - and the outer
oxygen atoms of MOE substituents are connected to phosphate
groups from 30 -adjacent residues via single- or double-water
bridges, a particular feature previously observed in the crystal
structures of 20 -O-MOE-modified RNAs (20,23) and shared
by other modifications with 20 -O-substituents containing
hydrogen-bond acceptors and/or donors (37). The presence
of a stable network of water molecules in the minor groove
that connects m5s2U bases with sugar moieties and phosphate
groups provides an indication that the 2-thio modification does
not significantly alter the water structure compared with uracil.
This observation alone does not allow conclusions as to the
role of hydration in the overall stabilizing effect of the 2-thio
modification. Nevertheless, the absence of obvious disruptions
in the minor groove hydration by sulfur atoms in the
m5s2Umoe duplex structure clearly argues against a destabilizing contribution.
DISCUSSION
The primary goal of the present study was an assessment of the
structural perturbations induced by the 2-thio modification in
an A-form duplex environment. The m5s2Umoe nucleoside
does not occur naturally; the choice to focus on this modification rather than the natural, biologically active s2U and m5s2U
pyrimidines (13,38) is based on the expectation that the 2-thio
modification alone will most probably not provide sufficient
nuclease resistance for in vivo and cell-based RNAi or antisense applications. A facile synthesis of the m5s2Umoe phosphoramidite building block for solid-phase synthesis has been
reported, and, indeed, the combination of the 2-thio and 20 -OMOE modifications confers increased nuclease resistance
based on in vitro measurements of resistance against snake
venom phosphodiesterase degradation (half-lives of oligos
with m5s2Umoe modifications >24 h) (11).
Replacement of the 2-oxygen of U or T by sulfur was shown
to greatly increase RNA affinity of base-modified oligonucleotides (9–11). The combination of the 2-thio and 20 -O-MOE
modifications also resulted in a higher thermodynamic stability of duplexes between modified oligonucleotides and complementary DNA (11), albeit to a lesser extent compared with
the corresponding RNAs. One would not expect a modification
5304
Nucleic Acids Research, 2005, Vol. 33, No. 16
Figure 7. Improvement of stacking interactions as a result of the 2-thio modification. Panels on the left show stacking interactions in the m5s2Umoe-modified duplex
and panels on the right show stacking interactions at the corresponding steps in the m5Umoe-modified reference duplex. Base-pair steps (A) (G3)p(T4):(A17)p(C18),
(B) (T4)p(A5):(m5s2Umoe16)p(A17), (C) (A5)p(m5s2Umoe6):(A15)p(m5s2Umoe16), (D) (m5s2Umoe6)p(A7):(T14)p(A15) and (E) (A7)p(C8):(G13)p(T14). G, C,
A and T are colored green, yellow, red and blue, respectively, and sulfur atoms are highlighted as yellow spheres. The figures were generated with the program
3DNA (36).
that brings about a considerable increase in the stability of
duplexes to significantly alter the geometry of the nucleobase,
sugar or phosphate portions, and/or the local and global conformations of a secondary or tertiary nucleic acid structural
motif. This assumption is based on observations from crystallographic data, indicating that stabilizing modifications commonly mimic the native conformation, whereas those that lead
to a loss of stability often display significant deviations,
Nucleic Acids Research, 2005, Vol. 33, No. 16
including sterically unfavorable interactions (20,39). Right
from the outset of the work described here, we were aware
that effects triggered by 2-thio modification would be rather
subtle and hence their detection would require relatively precise structural data. The resolution of the structure of the Aform decamer duplex with m5s2Umoe modifications, although
not extremely high, should be sufficient to ferret out some of
the consequences of the 2-thio and sugar modifications for
base-pair geometry and duplex conformation. Likewise, availability of a structure as a reference molecule that differs from
the duplex investigated here only by the absence of the sulfur
atom in the m5Umoe residues renders the analysis more
meaningful as it allows for a separation of various overlaid
effects.
Key features of the m5s2Umoe residues in the crystal structure of the modified DNA duplex are the C30 -endo pucker and
various subtle changes in the base pair, base-pair step and
base-pair helical parameters. In addition, the sulfur atoms
are in van der Waals contact across the minor groove as a
result of y- and x-displacements at the central base-pair step
(Figures 5 and 7C) and, judging from base overlaps in the
vicinity of m5s2Umoe residues, 2-thio modification leads to
small but measurable improvements in the stacking interactions compared with the reference structure. Among the
changes in base-pair geometry, the altered opening between
m5s2Umoe6 and A15 is readily apparent (Figures 5A and
6A). In fact, this is perhaps the most straightforward way to
accommodate the larger sulfur atom opposite the C2–H2 function of adenine. Increased propeller twist, stagger or shear
relative to a canonical geometry of the native T:A pair in
principle could each be used separately to relieve the steric
challenge posed by the larger sulfur. However, the opening
seen in the m5s2Umoe6:A15 pair is unique in the sense that this
adjustment seems sufficient alone to accommodate the sulfur
atom. In contrast, it appears that, rather than relying on a single
parameter to adjust to the bulkier sulfur, a combination of
adjustments including all of the above parameters is responsible for generating the particular arrangement seen in the
A5:m5s2Umoe16 pair.
Sugar pucker, nucleoside conformation and base-pair geometry are the result of a host of factors, including stereoelectronic effects. However, unlike improvements in stacking or
extensive hydration networks (the structure reveals both), the
structure does not allow any conclusions per se regarding
either a favorable or unfavorable contribution by some of
these factors to the net increase in stability observed with
the m5s2Umoe modification. In other words, although 2-thio
modification will probably cause a stronger preference of the
sugar for the C30 -endo pucker, all we see in the structure is the
product of the various stereoelectronic effects that bear on the
conformation and we can only speculate as to the relative
magnitude of the individual contributions toward the net
gain in stability. One of these (hidden) contributions is the
strengthening of the O40 –C10 –N1 anomeric effect in m5s2U
compared with thymine (see reference (40), pp. 43–46, for a
discussion of the relation between sugar pucker type and
molecular orbital overlap and hyperconjugation). Because
oxygen is a more electronegative element than sulfur, it
may appear counter-intuitive to consider sulfur a better
acceptor of negative charge than oxygen. Indeed, oxygen
has a greater ability to attract electron density per unit surface
5305
area. But if the greater surface area of sulfur compared with
oxygen is taken into account (factor 2.5), and the potentials for
each atom are summed over their respective spherical surfaces,
sulfur may well end up having an equal or greater potential
than oxygen. Another effect that could contribute favorably to
the higher stability afforded by the m5s2Umoe modification is
a change in Watson–Crick hydrogen bonding strength compared with RNA A:U or DNA A:T pairs. The N1. . .N3 hydrogen bond between A:U pairs in RNA was recently shown to be
stronger than the corresponding interaction in DNA (41).
With the m5s2Umoe nucleoside, the modulation of stereoelectronic effects extends into the 20 -O-substituent. For
example, the gauche effect between O40 and O20 that drives
the sugar conformational equilibrium toward the C30 -endo
side is enhanced by the presence of electronegative atoms
or groups in the substituent, as recently demonstrated by
extensive analyses of the influence of the chemistry of
RNA 20 -O-substituents on the stability of duplexes
(37,39,42–44). Combining the 2-thio and 20 -O-MOE modifications generates synergistic stereoelectronic effects that are
absent, e.g. in oligonucleotides modified by 20 -O-methyl-2thiouridine (10). These additive effects are the basis for the
high RNA affinity and increased nuclease resistance of the
m5s2Umoe modification.
Nature’s use of the s2U and m5s2U pyrimidines (i.e. in
tRNA) is clearly related to the conformational rigidity of
their ribose moiety as a result of the substitution of the 2oxygen by sulfur. The influence of the sulfur substitution is
demonstrated by the altered binding affinity and cleavage
activity of RNase H, an enzyme that is exquisitely sensitive
to conformational or steric changes in DNA:RNA substrates
(45). Although the enzyme binds RNA duplexes they are not
recognized as substrates (46), presumably because a canonical
A-form geometry precludes endonucleolytic cleavage of the
RNA strand. Consistent with this observation, incorporation of
m5s2U into the DNA strand of heteroduplexes dramatically
reduced the cleavage rate at ribonucleotides opposite modified
residues (47). Although this puts limitations on the use of this
modification in antisense applications, where RNase Hmediated degradation of the target mRNA is considered
important for efficacy, the properties of the s2U, m5s2U and
m5s2Umoe modifications may render them of interest for use
in siRNAs. Together with DNA phosphorothioates (8) and
40 -thio-RNA (3,48,49), they constitute the third class of
sulfur-modified nucleic acid building blocks with promising
features for the generation of nucleic acid therapeutics.
ACKNOWLEDGEMENTS
We are grateful to Dr Zdzislaw Wawrzak for help with data
collection and processing, and to Dr Pradeep S. Pallan for
comments on the manuscript. Use of the Advanced Photon
Source was supported by the US Department of Energy,
Office of Science, Office of Basic Energy Sciences, under
Contract No. W-31-109-Eng-38. The DuPont-NorthwesternDow Collaborative Access Team (DND-CAT) Synchrotron
Research Center at the Advanced Photon Source (Sector 5)
is supported by E. I. DuPont de Nemours & Co., The Dow
Chemical Company, the National Science Foundation, and
the State of Illinois. This work was supported by the US
National Institutes of Health (grant GM55237 to M.E.).
5306
Nucleic Acids Research, 2005, Vol. 33, No. 16
Funding to pay the Open Access publication charges for this
article was provided by grant NIH R01 GM55237.
Conflict of interest statement. None declared.
REFERENCES
1. Crooke,S.T. (1998) Basic principles of antisense therapeutics.
In Crooke,S.T. (ed.), Antisense Research and Application,
Vol. 131, Springer, Berlin, pp. 1–50.
2. Manoharan,M. (1999) 20 -Carbohydrate modifications in antisense
oligonucleotide therapy: importance of conformation, configuration and
conjugation. Biochim. Biophys. Acta., 1489, 117–130.
3. Manoharan,M. (2004) RNA interference and chemically modified small
interfering RNAs. Curr. Opin. Chem. Biol., 8, 570–579.
4. Soutschek,J., Akinc,A., Bramlage,B., Charisse,K., Constien,R.,
Donoghue,M., Elbashir,S., Geick,A., Hadwiger,P., Harborth,J. et al.
(2004) Therapeutic silencing of an endogenous gene by systemic
administration of modified siRNAs. Nature, 432, 173–178.
5. Egli,M. (1998) Towards the structure-based design of nucleic acid
therapeutics. In Weber,G. (ed.), Advan. Enzyme Regul.
Vol. 38, Elsevier Science Ltd., Oxford, UK, pp. 181–203.
6. Levin,A.A. (1999) A review of issues in the pharmacokinetics and
toxicology of phosphorothioate antisense oligonucleotides.
Biochim. Biophys. Acta., 1489, 69–84.
7. Filmore,D. (2004) Assessing antisense. Modern Drug Discovery
June, 49–50.
8. Crooke,S.T. (1995) Phosphorothioate oligonucleotides. In Crooke,S.T.
(ed.), Therapeutic Applications of Oligonucleotides. R. G. Landes,
Austin, pp. 63–79.
9. Kumar,R.K. and Davis,D.R. (1997) Synthesis and studies on the effect of
2-thiouridine and 4-thiouridine on sugar conformation and RNA duplex
stability. Nucleic Acids Res., 25, 1272–1280.
10. Shohda,K., Okamoto,I., Wada,T., Seio,K. and Sekine,M. (2000)
Synthesis and properties of 20 -O-methyl-2-thiouridine and
oligoribonucleotides containing 20 -O-methyl-2-thiouridine.
Bioorg. Med. Chem. Lett., 10, 1795–1798.
11. Rajeev,K.G., Prakash,T.P. and Manoharan,M. (2003) 20 -Modified-2thiothymidine oligonucleotides. Org. Lett., 5, 3005–3008.
12. Grosjean,H. and Benne,R. (eds) (1998) Modification and Editing of RNA.
ASM Press, Washington, DC.
13. Rozenski,J., Crain,P.F. and McCloskey,J.A. (1999) The RNA
modification database: 1999 update. Nucleic Acids Res.,
27, 196–197.
14. Agris,P. (2004) Decoding the genome: a modified view. Nucleic Acids
Res., 32, 223–238.
15. Testa,S.M., Disney,M.D., Turner,D.H. and Kierzek,R. (1999)
Thermodynamics of RNA–RNA duplexes with 2- or 4-thiouridines:
implications for antisense design and targeting of a group I intron.
Biochemistry, 38, 16655–16662.
16. Yokoyama,S., Watanabe,T., Murao,K., Ishikura,H., Yamaizumi,Z.,
Nishimura,S. and Miyazawa,T. (1985) Molecular mechanism of codon
recognition by tRNA species with modified uridine in the first position of
the anticodon. Proc. Natl Acad. Sci. USA, 82, 4905–4909.
17. Sierzputowska-Gracz,H., Sochaka,E., Malkiewicz,A., Kuo,K.,
Gehrke,C.W. and Agris,P.F. (1987) Chemistry and structure of modified
uridines in the anticodon, wobble position of transfer RNA are determined
by thiolation. J. Am. Chem. Soc., 109, 7171–7177.
18. Agris,P.F., Sierzputowska-Gracz,H., Smith,W., Malkiewicz,A.,
Sochacka,E. and Nawrot,B. (1992) Thiolation of uridine carbon-2
restricts the motional dynamics of the transfer RNA wobble position
nucleoside. J. Am. Chem. Soc., 114, 2652–2656.
19. Smith,W.S., Sierzputowska-Gracz,H., Sochacka,E., Malkiewicz,A. and
Agris,P. (1992) Chemistry and structure of modified uridine
dinucleosides are determined by thiolation. J. Am. Chem. Soc., 114,
7989–7997.
20. Tereshko,V., Portmann,S., Tay,E.C., Martin,P., Natt,F., Altmann,K.-H.
and Egli,M. (1998) Correlating structure and stability of DNA duplexes
with incorporated 20 -O-modified RNA analogues. Biochemistry, 37,
10626–10634.
21. De Mesmaeker,A., Häner,R., Martin,P. and Moser,H.E. (1995) Antisense
oligonucleotides. Acc. Chem. Res., 28, 366–374.
22. Martin,P. (1995) New access to 20 -O-alkylated ribonucleosides and
properties of 20 -O-alkylated oligoribonucleotides.
Helv. Chim. Acta., 78, 486–504.
23. Teplova,M., Minasov,G., Tereshko,V., Inamati,G.B., Cook,P.D.,
Manoharan,M. and Egli,M. (1999) Crystal structure and improved
antisense properties of 20 -O-(2-methoxyethyl)-RNA.
Nature Struct. Biol., 6, 535–539.
24. Berger,I., Kang,C.H., Sinha,N., Wolters,M. and Rich,A. (1996) A highly
efficient 24-condition matrix for the crystallization of nucleic acid
fragments. Acta Cryst. D, 52, 465–468.
25. Kabsch,W. (1993) Automatic processing of rotation diffraction data from
crystals of initially unknown symmetry and cell constants. J. Appl. Cryst.,
26, 795–800.
26. Navaza,J. (1997) AMoRe: an automated molecular replacement program
package. Meth. Enzymol., 276, 581–594.
27. CCP4. (1994), Collaborative computing project number 4. The CCP4
suite: programs for protein crystallography. Acta Cryst. D, 50, 760–763.
28. Egli,M., Tereshko,V., Teplova,M., Minasov,G., Joachimiak,A.,
Sanishvili,R., Weeks,C.M., Miller,R., Maier,M.A., An,H., Cook,P.D. and
Manoharan,M. (2000) X-ray crystallographic analysis of the hydration of
A- and B-form DNA at atomic resolution. Biopolymers (Nucleic Acid
Sciences), 48, 234–252.
29. Brünger,A.T., Adams,P.D., Clore,G.M., DeLano,W.L., Gros,P.,
Grosse-Kunstleve,R.W., Jiang,J.S., Kuszewski,J., Nilges,M.,
Pannu,N.S., Read,R.J., Rice,L.M., Simonson,T. and Warren,G.L. (1998)
Crystallography & NMR System: a new software suite for
macromolecular structure determination. Acta Cryst. D, 54, 905–921.
30. Brünger,A.T. (1992) Free R value: a novel statistical quantity for assessing
the accuracy of crystal structures. Nature, 355, 472–475.
31. Murshudov,G.N., Vagin,A.A. and Dodson,E.J. (1997) Refinement of
macromolecular structures by the maximum-likelihood method. Acta.
Cryst. D, 53, 240–255.
32. Murshudov,G.N., Vagin,A.A., Lebedev,A., Wilson,K.S. and Dodson,E.J.
(1999) Efficient anisotropic refinement of macromolecular structures
using FFT. Acta Cryst. D, 55, 247–255.
33. Winn,M.D., Isupov,M.N. and Murshudiv,G.N. (2001) Use of TLS
parameters to model anisotropic displacements in macromolecular
refinement. Acta. Cryst. D, 57, 122–133.
34. Egli,M., Usman,N. and Rich,A. (1993) Conformational influence of the
ribose 20 -hydroxyl group: crystal structures of DNA–RNA chimeric
duplexes. Biochemistry, 32, 3221–3237.
35. Tereshko,V., Wilds,C.J., Minasov,G., Prakash,T.P., Maier,M.A.,
Howard,A., Wawrzak,Z., Manoharan,M. and Egli,M. (2001) Detection of
alkali metal ions in DNA crystals using state-of-the-art X-ray diffraction
experiments. Nucleic Acids Res., 29, 1208–1215.
36. Lu,X-J., Shakked,Z. and Olson,W.K. (2000) A-DNA conformational
motifs in ligand-bound double helices. J. Mol. Biol., 300, 819–840.
37. Egli,M., Minasov,G., Tereshko,V., Pallan,P.S., Teplova,M.,
Inamati,G.B., Lesnik,E.A., Owens,S.R., Ross,B.S., Prakash,T.P. and
Manoharan,M. (2005) Probing the influence of stereoelectronic effects on
the biophysical properties of oligonucleotides: comprehensive analysis of
the RNA affinity, nuclease resistance and crystal structure of ten 20 -Oribonucleic acid modifications. Biochemistry, 44, 9045–9057.
38. Lagoja,I.M. (2005) Pyrimidine as a constituent of naturally biologically
active compounds. Chemistry & Biodiversity, 2, 1–50.
39. Pattanayek,R., Sethaphong,L., Pan,C., Prhavc,M., Prakash,T.P.,
Manoharan,M. and Egli,M. (2004) Structural rationalization of a large
difference in RNA affinity despite a small difference in chemistry between
two 20 -O-modified nucleic acid analogs. J. Am. Chem. Soc., 126,
15006–15007.
40. Thibaudeau,C. and Chattopadhyaya,J. (1999) Stereoelectronic Effects in
Nucleosides and Nucleotides and their Structural Implications. Uppsala
University Press, Uppsala, Sweden.
41. Vakonakis,I. and LiWang,A.C. (2004) N1. . .N3 hydrogen bonds of A:U
base pairs of RNA are stronger than those of A:T base pairs of DNA. J. Am.
Chem. Soc., 126, 5688–5689.
42. Prakash,T.P., Manoharan,M., Fraser,A.S., Kawasaki,A.M., Lesnik,E.,
Sioufi,N., Leeds,J.M., Teplova,M. and Egli,M. (2002) 20 -O-[2(Methylthio)ethyl]-modified oligonucleotide: an analog of 20 -O-[2(methoxy)ethyl]-modified oligonucleotide with improved protein
binding properties and high binding affinity to target RNA. Biochemistry,
41, 11642–11648.
43. Prhavc,M., Prakash,T.P., Minasov,G., Egli,M. and Manoharan,M. (2003)
20 -O-{2-[2-(N,N-Dimethylamino)ethoxy]ethyl} modified antisense
Nucleic Acids Research, 2005, Vol. 33, No. 16
oligonucleotides: symbiosis of charge interaction factors and
stereoelectronic effects. Org. Lett., 5, 2017–2020.
44. Prakash,T.P., Pueschl,A., Lesnik,E., Tereshko,V., Egli,M. and
Manoharan,M. (2004) 20 -O-[2-(guanidinium)ethyl]-5-methyluridine
modified oligonucleotides: synthesis and stabilizing effect on duplex and
triplex nucleic acid structures. Org. Lett., 6, 1971–1974.
45. Minasov,G., Teplova,M., Nielsen,P., Wengel,J. and Egli,M. (2000)
Structural basis of cleavage by RNase H of hybrids of arabinonucleic acids
and RNA. Biochemistry, 39, 3525–3532.
46. Lima,W.F. and Crooke,S.T. (1997) Binding affinity and specificity of
Escherichia coli RNase H1: impact on the kinetics of catalysis of antisense
oligonucleotide-RNA hybrids. Biochemistry, 36, 5004–5019.
5307
47. Lima,W.F., Nichols,J.G., Wu,H., Prakash,T.P., Migawa,M.T.,
Wyrzykiewicz,T.K., Bhat,B. and Crooke,S.T. (2004) Structural
requirements at the catalytic site of the heteroduplex substrate for human
RNase H1 catalysis. J. Biol. Chem., 279, 36317–36326.
48. Hoshika,S., Minakawa,N. and Matsuda,A. (2004) Synthesis and physical
and physiological properties of 40 -thioRNA: application to postmodification of RNA aptamer toward NF-kB. Nucleic Acids Res.,
32, 3815–3825.
49. Haeberli,P., Berger,I., Pallan,P.S. and Egli,M. (2005) Syntheses of 40 thioribonucleosides and thermodynamic stability and crystal structure of
RNA oligomers with incorporated 40 -thiocytosine. Nucleic Acids Res.,
33, 3965–3975.